# Ghana Education Analytics
This project presents an exploratory analysis of anonymized continuous assessment
data from a Ghanaian public basic school. The study demonstrates a reproducible
education analytics workflow and investigates how student performance varies
across subjects and assessment types.
## Study Context
The data were collected from a public basic school in Ghana serving upper primary
students (Basic Eight). The school follows the national curriculum prescribed by
the Ghana Education Service (GES) and serves learners from mixed socioeconomic
backgrounds. To protect confidentiality, no identifying information about the
school or students is disclosed.
A detailed description of the study setting, student population, and assessment
structure is provided in `docs/study_context.md`.
## Research Questions
1. How does student performance vary across subjects?
2. How does performance differ by assessment type?
3. Which subjects exhibit high variability, indicating potential learning gaps?
4. How do subjects respond differently to continuous assessment and formal
examinations?
## Data
- **Raw data**: Microsoft Excel files containing student assessment records
(not included for ethical reasons).
- **Processed data**: Fully anonymized, long-format dataset used for analysis
(`data/processed/assessment_long.csv`).
## Methodology
1. Data auditing and ethical anonymization
2. Reshaping multi-sheet assessment data into tidy long format
3. Exploratory data analysis across subjects and assessment types
## Repository Structure
ghana-education-analytics/
├── notebooks/ # Jupyter notebooks for each project phase
├── data/
│ ├── raw/ # Placeholder only (raw data not shared)
│ └── processed/ # Anonymized datasets
├── docs/ # Study context and documentation
└── requirements.txt # Python dependencies
## Tools
- Python
- pandas
- matplotlib
- Jupyter Notebook
## Ethics
All student identifiers were removed prior to analysis. The project uses only
anonymized data and i …